The step of the machine learning process that involves making predictions using the trained model is prediction, also sometimes referred to as inference. This is the final step in the machine learning pipeline and the reason why we train models in the first place.
Here's why prediction is a separate step:
- Separation of concerns: Training and prediction involve different processes and goals. Training focuses on learning a model from the data, while prediction focuses on applying the learned model to make new predictions. Separating these steps ensures that the model is not overfitting to the training data and can generalize to unseen data.
- Computational efficiency: Once a model is trained, it can be used to make predictions on new data much faster than training the model from scratch each time. This is especially important for large datasets or real-time applications.
- Evaluation: Prediction allows us to evaluate the performance of the trained model on unseen data. This is crucial for assessing the effectiveness of the model and identifying areas for improvement.
In summary, prediction is a distinct step in the machine learning process because it allows for efficient and accurate application of the trained model, facilitates model evaluation, and ensures generalization to unseen data. This is the stage where we finally reap the benefits of the entire machine learning process, applying the learned knowledge to solve real-world problems.
